Forum Discussion

MangoMagic's avatar
MangoMagic
Regular Visitor
1 year ago
Solved

VS Code Fabric User Data Functions

I was trying to develop Fabric User Data Functions on my local machine as doing it through Fabric portal is very difficult and it takes many minutes to publish / test & debug function after making mi...
User Data Functions
  • v-dineshya's avatar
    v-dineshya
    1 year ago

    Hi MangoMagic ,

    Thank you for reaching out to the Microsoft Community Forum.

     

    You are trying to import "fabric.functions", which is not part of the open-source fabric PyPI package. In Microsoft Fabric, User Defined Functions (UDFs) for the Data Factory or Lakehouse experience are powered through a special runtime in the Fabric service not a publicly installable Python package like fabric.functions.

     

    So when you try to import

     

    import fabric.functions as fn

     

    You are assuming that "fabric.functions" is an importable local module but it's only available in the Microsoft Fabric runtime environment, not on your local machine.


    1. Microsoft Fabric UDFs rely on a specific SDK that is not publicly available via PyPI. It's bundled within the Fabric runtime environment. You cannot run or test these functions locally using the standard Python environment.You must use the Fabric portal or a supported development environment that includes the Fabric runtime.

     

    2. You can test the fabric.functions module locally by creating a sample module with the same interface for testing purposes. Alternatively, use logging and unit tests to validate your logic independently of the Fabric-specific decorators.

     

    Please refer Microsoft document

    Microsoft Fabric documentation - Microsoft Fabric | Microsoft Learn

     

    I hope this information helps. Please do let us know if you have any further queries.

     

    Regards,

    Dinesh

v-dineshya's avatar
v-dineshya
Community Support
1 year ago

Hi MangoMagic ,

Thank you for reaching out to the Microsoft Community Forum.

 

You are trying to import "fabric.functions", which is not part of the open-source fabric PyPI package. In Microsoft Fabric, User Defined Functions (UDFs) for the Data Factory or Lakehouse experience are powered through a special runtime in the Fabric service not a publicly installable Python package like fabric.functions.

 

So when you try to import

 

import fabric.functions as fn

 

You are assuming that "fabric.functions" is an importable local module but it's only available in the Microsoft Fabric runtime environment, not on your local machine.


1. Microsoft Fabric UDFs rely on a specific SDK that is not publicly available via PyPI. It's bundled within the Fabric runtime environment. You cannot run or test these functions locally using the standard Python environment.You must use the Fabric portal or a supported development environment that includes the Fabric runtime.

 

2. You can test the fabric.functions module locally by creating a sample module with the same interface for testing purposes. Alternatively, use logging and unit tests to validate your logic independently of the Fabric-specific decorators.

 

Please refer Microsoft document

Microsoft Fabric documentation - Microsoft Fabric | Microsoft Learn

 

I hope this information helps. Please do let us know if you have any further queries.

 

Regards,

Dinesh

Anonymous's avatar
Anonymous
Not applicable
8 months ago

I have the same issue. But the difference is that I am trying to do this on the Fabric UI directly and I do get the same "ModuleNotFoundError" even though the fabric-user-data-functions is part of my environment

  • v-dineshya's avatar
    v-dineshya
    Community Support
    8 months ago

    Hi Anonymous ,

    Thank you for reaching out to the Microsoft Community Forum.

     

    Please try below steps to fix the issue.

     

    1. Re-publish the UDF item,

    Open the UDF item --> Library management. Confirm fabric_user_data_functions is present, it should be and is non-removable. Click Publish to ensure the runtime actually picks the listed versions.


    2. Replace your code temporarily with the below sample to isolate the error:

     

    from datetime import datetime
    import logging
    import fabric.functions as fn

    udf = fn.UserDataFunctions()

    @udf.function()
    def hello_fabric(name: str) -> str:
    logging.info("Python UDF trigger function processed a request.")
    return f"Welcome to Fabric Functions, {name}, at {datetime.now()}!"

     

    Note: if this fails, it’s not your logic. it’s the runtime/library binding.

     

    3. Run the function from portal or an invocation and then open View logs for the UDF item. Look for the exact stack and the runtime version being used. UDF doc set shows the logging approach; portal has a logs view for UDF runs.

     

    4. Remove any conflicting library entries, If you previously added fabric the SSH/task-runner package at the UDF item or workspace, remove it. it doesn’t contain fabric.functions and can cause confusion. The correct SDK is fabric_user_data_functions.

     

    5. Recreate the UDF item, from VS Code (Fabric Workspace explorer), create a new UDF item, choose Python 3.11, open locally, and publish. This resets the item with a clean configuration and the right SDK. The Quickstart/Marketplace pages document the exact flow, including local run and publish.

     

    Please refer below links.

    Quickstart - Create a Fabric User data functions item in Visual Studio Code - Microsoft Fabric | Microsoft Learn

    fabric.functions package | Microsoft Learn

    Python programming model for Fabric User data functions - Microsoft Fabric | Microsoft Learn

    How to manage libraries for your Fabric User Data Functions - Microsoft Fabric | Microsoft Learn

     

    I hope this information helps. Please do let us know if you have any further queries.

     

    Regards,

    Dinesh